Applied Scientist, Core Search

About the role

Amazon Core Search research role spanning reasoning LLMs, embeddings, rankers, and multi-objective optimization; includes LLM mid/post-training, RLHF/DPO, contrastive learning, and large-scale experiments.

Requirements

• PhD, or Master's plus 4+ years in CS, computer engineering, ML, or related field
• Programming in Python, Java, C++, or related language
• Experience in algorithms, parsing, optimization, data mining, parallel/distributed computing, or HPC
• Professional software development experience preferred

What you'll do

• Develop personalized multimodal reasoning LLMs for queries and products
• Mid-train and post-train LLMs, including continued pretraining and instruction tuning
• Align models with RLHF, preference optimization, and reward models
• Build product and customer representations with embeddings and contrastive learning
• Develop LLM rankers and multi-objective ranking systems
• Improve offline benchmarks and online experiments

Where this was posted

• LinkedIn

  last seen 2h ago
  View on LinkedIn